Analysis of Sequential Organ Failure Assessment (SOFA) Score

Profile in Relation to Length of Stay and Patient Outcome in the ICU

of Abdul Moeloek Regional General Hospital

Ari Wahyuni

1,*

, Liana Sidharti

and Desy Kusumaningrum

Department of Anesthesiology and Intensive Care, Faculty of Medicine, University of Lampung,

Bandar Lampung, Indonesia

Faculty of Medicine,University of Lampung, Bandar Lampung, Indonesia

Keywords: ICU, Length of Stay, Mortality, Sepsis, SOFA Score.

Abstract: Sepsis is a severe dysfunction of the body's organs triggered by the immune system's imbalance in reaction to

an infection. The Sequential Organ Failure Assessment (SOFA) score is utilized to evaluate the organ failure

linked to sepsis. An increase in the SOFA score is associated with worse patient outcomes or higher mortality

rates. This study seeks to analyze the correlation between SOFA scores, length of stay, and mortality in sepsis

patients. The study subjects are sepsis patients treated in the Intensive Care Unit (ICU) of Dr. H. Abdul

Moeloek Regional General Hospital. The research was conducted from May to October 2023. This study is a

prospective cohort analytical observational research conducted at Dr. H. Abdul Moeloek Regional General

Hospital and the University of Lampung between May and October 2023. A total of 137 patients were

involved during a span of three months, with 89 patients chosen as subjects for this study, while 48 patients

were excluded because of incomplete data. The Spearman Correlation normality test displayed a correlation

between the SOFA score and Length of Stay with a significance value of 0.367. Since this value is >0.05, H0

is accepted, and H1 is rejected. The Spearman Correlation normality test indicated a correlation between the

SOFA score and the patients' final condition with a significance value of 0.000. As this value is <0.05, H0 is

rejected, and H1 is accepted. The correlation coefficient of 0.097 indicates a very weak level of association

between the SOFA score and Length of Stay, suggesting no significant correlation between the two variables.

However, a correlation coefficient of 0.592 denotes a moderately robust level of association between the

SOFA score and the Patients' Final Condition, highlighting a correlation between the two variables.

1 INTRODUCTION

Sepsis is a health issue prevalent in society

characterized by a severe clinical infection syndrome

marked by cardinal inflammatory signs such as

leukocyte accumulation, vasodilation, and increased

microvascular permeability occurring in tissues far

from the source of infection (Darwis & Probosuseno,

2019).

According to the WHO (2017), there were 48.9

million reported cases with 11 million deaths

associated with sepsis, contributing for 20% of all

deaths worldwide. Each year, sepsis causes 6 million

deaths and has been designated as a global health

priority by the WHO. Sepsis cases rank among the top

ten causes of death in the United States (Marik &

Taeb, 2011).

A study conducted in 2009 across 16 Asian

countries, including Indonesia, stated that the

incidence of severe sepsis and septic shock in ICU

wards was 10.9%, with a corresponding mortality rate

of 44 percent. Another study at RSCM Jakarta in

2012, spanning one month, it was found that among

84 intensive care cases, there were 23 instances of

severe sepsis and septic shock, resulting in a mortality

rate of 47.8 percent (Kemenkes, 2017).

Based on data retrieved from patient medical

records at RSUP Dr. Sardjito Yogyakarta in 2016, the

incidence of sepsis in internal disease inpatient wards

amounted to 704 cases (17.06%) out of the total

number of treated patients, with 431 (61.22%) deaths

among diagnosed sepsis patients. In 2017, there was

a decrease in sepsis incidence to 454 cases (9.71%)

out of the total number of treated patients, with 278

Wahyuni, A., Sidharti, L. and Kusumaningrum, D.

Analysis of Sequential Organ Failure Assessment (SOFA) Score Proﬁle in Relation to Length of Stay and Patient Outcome in the ICU of Abdul Moeloek Regional General Hospital.

DOI: 10.5220/0013668100003873

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Medical Science and Health (ICOMESH 2023), pages 203-207

ISBN: 978-989-758-740-5

203

(61.23%) deaths among diagnosed sepsis patients

(Darwis & Probosuseno, 2019).

Assessment systems for organ damage and failure

can be monitored using ICU illness scores such as

APACHE II, SAPS II, and the SOFA score to assess

the severity of the disease (Dirgantoro, 2018). The

APACHE II scoring system itself has several

limitations due to selection bias, lead time bias, and

the difficulty in determining the primary diagnosis

leading patients into the ICU, requiring a higher cost

due to multiple laboratory variables tested and a

lengther time to obtain results. Whereas the SAPS II

system, although the first system to use statistical

modeling techniques, is difficult to establish

considering data must be collected within one hour

after the patient enters the ICU (Sugiman, 2011).

In sepsis, organ dysfunction can be recognized by

a sudden increase of at least 2 points in the overall

SOFA score because of an infection. Each organ has

a value ranging from 0 (normal function) to 4 (very

abnormal), resulting in a maximum total score of 24

(Seymour et al., 2016). The researchers chose the

SOFA score because it's not a one-time assessment; it

can be evaluated periodically, observing score

increases or decreases. The SOFA score isn't just

calculated upon patient admission but also every 24

hours. Parameters in the SOFA score are deemed

ideal for depicting organ dysfunction (Singer et al.,

2016). Moreover, the SOFA score attributes values to

individual organ systems depending on one or

multiple variables, fewer than other ICU severity

assessment systems like APACHE II and SAPS II

(Dirgantoro, 2018).

Identifying sepsis before significant organ failure

occurs poses a challenge to all medical professionals.

For primary care doctors, detecting potentially septic

patients among many individuals presenting with

uncomplicated infections, as well as those not

progressing to sepsis, requires specific skills (Tavare

& Oflyn, 2017). The lack of diagnostic tools is a

major obstacle in the early management of sepsis.

This is compounded by the fact that sepsis presents as

a heterogeneous set of symptoms without a gold

standard for diagnosis (Mclymont & Glover, 2016).

2 METHODS

2.1 Research Design

This research is an analytical observational research

using a prospective cohort study approach. It was

carried out at RSUD Dr. H. Abdul Moeloek and

University of Lampung from Mei to October 2023.

2.2 Sample Research

The subjects of this study were individuals admitted

to the ICU at Dr. H. Abdul Moeloek Regional General

Hospital who met the study criteria. Inclusion criteria

for this research were patients aged ≥18 years

diagnosed with either sepsis or septic shock.

Exclusion criteria included patients discharged

without the approval of the attending physician or

upon their own request, and patients with a SOFA

score <2 assessed based on serum creatinine

concentration, platelet count, and Glasgow Coma

Scale (GCS) within 24 hours following the diagnosis

of sepsis. A total of 137 patients were identified over

three months, with 89 patients included as subjects in

this study, while 48 patients were excluded due to

incomplete patient data.

2.3 Data Analysis Research

In this research, the data was performed on the data

using SPSS version 22.0. Spearman's test was utilized

for numerical data to evaluate the distribution of the

data. Descriptive data were reported as the mean and

standard deviation (SD) for normally distributed data

or as the median and interquartile range for data that

did not follow a normal distribution. Bivariate

analysis was conducted to compare variable

differences between subjects who experienced

improvement and those who deceased.

3 RESULTS

The SOFA score is one of the scoring systems

commonly used to depict organ failure or

dysfunction, typically measured in patients

undergoing treatment in the intensive care unit. The

influence distribution factor are presented in the table

Table 1: Distribution Frequency of Patients Age.

Data Criteria Patients Age

N 89

Minimun 18

Maximu

Mean 56.49

ICOMESH 2023 - INTERNATIONAL CONFERENCE ON MEDICAL SCIENCE AND HEALTH

204

Based on the research data provided, it's evident

that the ICU at Dr. H. Abdul Moeloek Regional

General Hospital had a total of 89 patients as research

subjects. The patients' ages displayed a considerable

range, spanning from a minimum of 18 years to a

maximum of 94 years. The average age of patients

treated in the ICU is approximately 56.49 years.

These results depict the diversity of ages among

patients requiring intensive care in this hospital.

Table 2: Distribution Frequency of Patients Gender.

Data Criteria Patients Gende

Men 42

Women 47

Total 89

Based on the research data provided, it is

observed that in the ICU of Dr. H. Abdul Moeloek

Regional General Hospital, out of a total of 89

research subjects, there were 42 male patients and 47

female patients. These results depict an insignificant

difference between the number of male and female

patients.

Table 3: Distribution Frequency of Primer Diagnoses.

Data Criteria Primer Diagnoses

Neurological Disorders 19

Cardiovascular System 15

Res

iration S

ste

Gastrointestinal S

stem 2

Endocrine Disorders 4

Infection 1

Postoperative Management 25

Liver Disorders 3

Kidne

Disorders 6

Mali

nanc

Sensory Syte

Hematological Disorders 2

Reproduction Syste

Total 89

Based on the distribution of primary diagnoses,

it was found that out of the total 89 research subjects,

the three most common primary diagnoses were

patients undergoing postoperative care, neurological

system disorders, followed by cardiovascular system

issues. Patients undergoing postoperative care

included those after a laparotomy. The most prevalent

neurological diseases were patients with brain

hemorrhage and stroke. As for cardiovascular

diseases, the most frequently encountered were

patients with heart failure and coronary artery

disease.

Table 4: Distribution Frequency of Number of

Comorbidities.

Data Criteria Number of Comorbidities

0 17

1 21

2 18

4 18

Total 89

Based on the distribution of data on the number

of patient comorbidities, it was found that the

majority of patients undergoing treatment in the ICU

at Dr. H. Abdul Moeloek Regional General Hospital

had comorbidities, varying from a minimum of 1 to a

maximum of 6 comorbidities. Patients without

comorbidities amounted to 17 individuals. The

presence of comorbidities in patients, whether

acquired before or during their treatment, becomes a

factor that worsens the patient's condition and poses a

higher risk of mortality. The most commonly found

comorbidities were electrolyte imbalance and

infections.

Table 5: Distribution Frequency of SOFA Score.

Data Criteria SOFA Score

N 89

Minimu

Maximu

Mean 7.44

Based on table 5, the Sequential Organ Failure

Assessment (SOFA) Score among patients varied

between a minimum score of 3 and a maximum score

of 14. The mean SOFA score was around 7.44. A

higher SOFA score indicates a greater severity of

organ dysfunction experienced by the patient.

Table 6: Distribution Frequency of Length Hospitalization.

Data Criteria Len

th Hos

italization

N 89

Minimu

Maximu

Mean 5.61

According to table 6, the length of stay varied

between 0 days as the minimum and 25 days as the

maximum. The average length of stay in this ICU was

approximately 5.61 days. The varied length of stay

from 0 to 25 days reflects the diversity in patient

conditions, with some patients possibly requiring

shorter treatment durations while others may need

lengther care.

Analysis of Sequential Organ Failure Assessment (SOFA) Score Proﬁle in Relation to Length of Stay and Patient Outcome in the ICU of

Abdul Moeloek Regional General Hospital

205

Table 7: Distribution Frequency of Final Condition.

Final Condition Fre

uenc

Percenta

(

)

Life 12 13.5

Dea

77 86.5

Total 89 100

Based on the above table 4, out of the total 89

patients who were subjects of the study,

approximately 13.5% of them successfully recovered

and survived after undergoing treatment in the ICU.

The most notable data is that around 86.5% of

patients experienced death during or after ICU

treatment.

Table 8: SOFA Score Normality Test Results with Length

of Hospitalization.

Variable P-Value

SOFA 0.008

Len

th Hos

italization 0.000

Based on the normality test using Kolmogorov-

Smirnov in the table above, it is known that the

probability values (p-values) or Asymp. Sig. (2-

tailed) for the variables SOFA Score and Length of

Stay are 0.008 and 0.000, respectively. As the

probability values are smaller than the significance

level of 0.05, it indicates that the normality

assumption is not met. Therefore, the correlation test

between SOFA Score and Length of Stay will be

conducted using Spearman Correlation.

Table 9: Correlation Test Results of SOFA Score with

Length of Hospitalization.

Variable Person (r) P-Value

SOFA

0.097

0.367

Length Hospitalization

From the table 9, it is revealed that concerning the

relationship between the SOFA Score and Length of

Stay yielded a significance value of 0.367. Since this

value is >0.05, H0 is accepted, and H1 is rejected,

indicating no relationship between the SOFA Score

and Length of Stay. With a correlation coefficient of

0.097, it signifies a very low level of association

(correlation) between the SOFA Score and Length of

Stay.

Table 10: SOFA Score Normality Test Results with

Patient’s Final Condition.

Variable P-Value

SOFA 0.008

Final Condition 0.000

Based on the normality test using Kolmogorov-

Smirnov in the table above, it is noted that the

probability values (p-values) or Asymp. Sig. (2-

tailed) for the variables SOFA Score and Patient

Outcome are 0.008 and 0.000, respectively. As the

probability values are smaller than the significance

level of 0.05, it indicates that the normality

assumption is not met. Therefore, the correlation test

between the SOFA Score and Patient Outcome will

be conducted using Spearman Correlation.

Table 11: Correlation Test Results of SOFA Score with

Patient’s Final Condition.

Variable Pearson

(r)

P-Value

SOFA

0.592

0.000

Kondisi Akhir Pasien

From the table 11, it's clear that in relation to the

connection between the SOFA Score and the patient's

outcome, a significance value of 0.000 was obtained.

As this value is <0.05, H0 is declined, and H1 is

acknowledged, signifying an association between the

SOFA Score and the Patient's Outcome. With a

correlation coefficient of 0.592, it signifies a

moderately strong level of association (correlation)

between the SOFA Score and the Patient's Outcome.

As the correlation coefficient is positive, it indicates

a direct relationship between the two variables.

Therefore, it can be concluded that as the SOFA

Score increases, the patient's outcome deteriorates.

4 DISCUSSION

The SOFA score is a reliable scoring method utilized

to illustrate organ failure or dysfunction, typically

measured in patients undergoing intensive care unit

(ICU) treatment. However, based on the findings of

this research, there was no correlation found between

the SOFA score and the duration of stay. The

correlation test revealed a weak correlation between

these two variables. Factors influencing ICU patient

length of stay include age, primary diagnosis, and the

patient's initial condition upon admission. In this

study, the subjects encompassed all patients receiving

ICU care, resulting in diverse patient ages and

primary diagnoses, hence no significant relationships

were found due to the variability in patient

backgrounds. This highlights the need for more

specific research in subsequent studies.

The SOFA score serves as an indication of the

quantity and severity of organ dysfunction in ICU

patients. A higher SOFA score indicates increased

organ dysfunction severity, elevating the risk of

mortality (Sari et al., 2021). SOFA score

ICOMESH 2023 - INTERNATIONAL CONFERENCE ON MEDICAL SCIENCE AND HEALTH

206

measurements indicated that the majority of patients

with a fatal outcome experienced multi-organ

dysfunction, notably in the respiratory system, as

observed from PaO2/FiO2 data. Patients

experiencing respiratory distress exhibited decreased

lung compliance and hypoxemia, leading to

inadequate oxygen reaching body tissues.

Another indicator of multi-organ dysfunction seen

in SOFA score measurements is the neurological

system through a decrease in the Glasgow Coma

Scale (GCS), signifying reduced patient

consciousness levels. This serves as an indication of

potential brain injury worsening due to infection,

masses, or other inflammatory processes. The

cardiovascular system is also assessable via the

SOFA score, with Mean Arterial Pressure (MAP) as

an indicator. The initial MAP target for patients

receiving vasopressor therapy is 65 mmHg to achieve

optimal tissue perfusion. Prolengthed hypotension is

associated with increased patient mortality risk.

Serum creatinine levels serve as an indicator for

monitoring kidney function and are included in SOFA

score evaluations. Patients with acute kidney injury

exhibit increased serum creatinine levels, indicating

weakened filtration and elimination over several

hours to days. Elevated serum creatinine levels

correlate with worse patient outcomes.

The research findings indicate a connection

between the SOFA score and the final condition of

ICU patients. Correlation tests showed that higher

SOFA scores correspond to an increased risk of the

patient's final condition worsening. This aligns with

previous studies where a SOFA score ≥7 had a

mortality rate of 72.6% (Sari et al., 2021). Iskandar

and Siska (2020) mentioned that individuals having a

SOFA score of ≥7 faced a mortality risk 2.8 times

higher than those with a SOFA score of <7. Other

studies, such as Bale et al. (2013), also highlighted the

use of the SOFA score as a predictor of the patient's

final condition.

5 CONCLUSION

The SOFA score does not have a significant

relationship with the length of stay of patients

undergoing intensive care treatment in the ICU at

RSUD Dr. H. Abdul Moeloek. However, the SOFA

score does have a significant relationship with patient

mortality, thus serving as a reasonably good predictor

in predicting the patient's final condition.

REFERENCES

Ascharya, S., Pradhan, B., and Marhatta, M., 2007.

Application of the Sequential Organ Failure

Assessment (SOFA) Score in Predicting Outcome in

ICU Patients with SIRS. Kathmandu University

Medical Journal. Vol.5, No.4, pp.475-483.

Darwis I amd Probosuseno., 2019. Hubungan Neutorphyl

Lymphocyte dengan Outcome Pasien Sepsis pada

Geriatri. JK Unila. Vol.3, No.1, pp.147-153.

Dirgantoro, Z., 2018. Tesis : Hubungan antara Red Cell

Distribution Width (RDW), neutrofil-Limfosit Rasio

(NLR), Mean Platelet Volume (MPV) dengan skor

SOFA sebagai Prediktor Keparahan pada Sepsis di

RSUD Dr. Moewardi Surakarta. Surakarta :

digilib.uac.id.

Ferreira, F., Giuseppe, D., Giovanni, B., Fransescoc, D.,

and Pasquale, P., 2017. Sepsis dan Septic Shock: New

Definitions, New Diagnostic and Therapeutic

Approaches. Journal of Global Antimicrobial

Resistance. Vol.10, pp.204-212.

Iskandar A and Siska F., 2020. Analisis Hubungan SOFA

Score dengan Mortalitas Pasien Sepsis. Jurnal

Kesehatan Andalas. Vol. 9, No.2, pp.168-173.

Kemenkes., 2017. Pedoman Nasional Pelayanan

Kedokteran Tata Laksana Sepsis. Jakarta : Kementerian

Kesehatan RI.

Marik P dan Taeb A., 2017. SIRS, qSOFA, and new sepsis

definition. J Thorac Dis. Vol.9, No.4, pp.943-945.

McLymont dan Glover G., 2016. Scoring systems for the

characterization of sepsis and associated outcomes.

Ann TransI Med. Vol.4, No.24, pp.527.

Sari, E., Hayati, Y., and Rokhmawati, N., 2021. Hubungan

Skor Sofa dengan Mortalitas pada Pasien Sakit Kritis.

Majalah Kesehatan. Vol.8, No.3, p.149-155.

Seymourl, W., Vincent, C., Theodore, J., Frank, M,,

Thomas, D., et al., 2016. Assessment of clinical criteria

for sepsis : For the third international consensus

definitions for sepsis and septic shock (Sepsis-3).

JAMA. Vol.315, pp.762-774.

Shapiro, N., Zimmer, G., and Barkin, A., 2010. Sepsis

Syndromes. In: Marx et al. Rosen’s Emergency

Medicine Concepts and Clinical Practice. 7th Ed.

Philadelphia : Mosby Elsevier.

Singer, M., Deutschman, C., Seymour, C., Shankar, H.,

Annanne D, et al., 2016. The Third International

Consensus Definitions for Sepsis and Septic Shock

(Sepsis-3). JAMA. Vol.315, No.8, pp.801-810.

Sugiman, T., 2011. Sistem Skor di Intensive Care Unit.

Majalah Kedokteran Terapi Intensif. Vol.1, No.2,

pp.76-88.

Tavare A and Oflynn N., 2017. Recognition, Diagnosis, and

Early Management of Sepsis : NICE Guideline. British

Journal of General Practice. Vol.67, pp.185-186.

WHO., 2017. Improving the Prevention, Diagnosis, and

Clinical Management of Sepsis. World Health

Organizations.

Analysis of Sequential Organ Failure Assessment (SOFA) Score Proﬁle in Relation to Length of Stay and Patient Outcome in the ICU of

Abdul Moeloek Regional General Hospital

207